WO1995030009A2 - Identification, production and use of saponin glyclosyl hydrolases - Google Patents

Abstract

Polypeptides with saponin glycosyl hydrolase activity, including avenacinase and tomatinase, have been purified and encoding nucleic acid therefor cloned and manipulated. The nucleic acid is useful in production of the polypeptides, for example in host cells such as plant or microbial cells, and in screening for other saponin glycosyl hydrolase genes. Oligonucleotides for use in probing have been designed using sequence information obtained. The polypeptides may be used in deglycosylating saponins, in raising antibodies which may be used in screening for other saponin glycosyl hydrolases, and in identifying molecules able to modulate saponin glycosyl hydrolase activity, which molecules are themselves useful. Polypeptides with saponin glycosyl hydrolase activity, and encoding nucleic acid therefor, have been obtained from Gaeumannomyces graminis var. tritici and graminis.

Description

IDENTIFICATION, PRODUCTION AND USE OF SAPONIN

GLYCOSYL HYDROLASES

The present invention relates to the identification, production and use of saponin glycosyl hydrolases, including the cloning of nucleic acid with sequences encoding these enzymes .

Saponins are glycosides occurring primarily but not exclusively in plants, and which have the property of forming a soapy lather when shaken with water (1,2) . The ability to cause frothing has been used to identify the presence of saponins in plants and plant extracts. Saponins, in general, are very powerful emulsifiers and are toxic, hemolytic and able to form complexes with cholesterol and other sterols which contain a 3 /3-hydroxyl group (2,3) . The formation of saponin/sterol complexes in the eukaryotic membranes (of fungi, plants and animals) results in pore formation and leakage of cell contents, with subsequent cell death (4,5) . Some properties generally ascribed to saponins, for example, bitterness, are better seen as characterizing particular types of saponin, rather than being shared by all members. Some saponins also have beneficial properties, which include lowering cholesterol levels (5,6,7) , anti- tumour (9) and anti-leukaemic (9) activity, therapeutic effects on metabolism (7,9) , and anti-rheumatic (7) and anti-inflammatory (6,7) activity. Saponins have been identified as the active components which confer the beneficial effects of ginseng, and are used in the treatment of cardiovascular disorders and as adjuvants (6,7) .

The antimicrobial activity of saponins has been well documented (1,2,6,7,8,10) and there is considerable interest in the use of saponins to treat causal agents of animal diseases (6,7) and also in the significance of saponins in conferring resistance of plants to attack by microbes and invertebrate pests including insects (1,7,8,10) . Saponins also serve as raw material for the industrial production of hormones (8) , and are valuable research tools for permeabilising membranes (1) . Thus, from the biological standpoint in its widest context, saponins possess a diversity of properties, some of which are deleterious, but many of which are beneficial. For this reason there has been great interest in the chemical characterisation of saponins responsible for the effect of plant drugs (e.g. digitoxin) and folk medicines (e.g. Ginseng, ginger and liquorice) , and more recently in food science and nutritional research.

Saponins are glycosides with structures based on one of the four aglycones shown in Figure 1. Prominent members of the different groups are: triterpenes - avenacin (Avena sativa) , cyclamin (Cyclamen persicum) , Qj-hederin (Hedera helix) ; alkaloid α-tomatine (Lycopersicon esculentum) ; α-solanine, α-chaconin ( Solanu tuberosum) ; spirostanol - digitonin (Digi talis purpurea) ; furostanol - avenacoside (Avena sativa) (ref. 8) . Saponins are quite stable compounds: for instance, the levels of α-chaconin and α-solanine in potatoes are unaffected by boiling, baking or microwaving (1) .

The alkaloids which occur in, for example, Solanaceous species (many of which are food crops) , have received considerable attention due to their toxicity (1) . In potatoes, high starch and glycoalkaloid contents appear to be closely genetically linked, making potato varieties bred for the starch industry unsuitable for human consumption. Saponins in tomato seeds are the cause for the bitter taste, making the debris from the tomato processing industry less attractive for use as animal feed.

Saponins in general are widely distributed in plant species, being reported in nearly 100 families (1) . Table 1 lists a selection of plants used as foods or feed stuffs that contain saponins. Saponins from some plants are highly toxic to man and other species (9) and are the cause of bitterness in food and feed crops such as, for example, cucurbits, horse chestnut, tomato seeds, sugar beets, soybeans and other legumes. Saponins from alfalfa are thought to play a role in causing flatulence.

The glycone parts of these compounds are generally oligosaccharide, linear or branched, attached to a hydroxyl or a carboxyl group or both. There may be one (monodesmosides) , two (bidesmosides) or three (tridesmosides) sites of attachment (6) . Those saponins which are bi- or tridesmodic are commonly present in plant tissue in a biologically inactive form, and are converted into active monodesmosidic forms by plant enzymes in response to tissue damage (6,9,10) . Monodesmosidic saponins commonly have much greater hemolytic and antimicrobial properties than their bi- or tridesmodic counterparts. For instance, intact oat leaves contain the inactive bisdesmosidic saponins known as avenacosides A and B (Figure 2) . Upon wounding of the cells, these are transformed into antifungal monodesmosidic 26-desglucoderivatives. The plant S-glycosidase catalysing this reaction, the 26-desglyco-avenacosidase, is highly specific since it splits off only the glucose molecule attached to C-26, but does not affect the glucose molecules of the side chain attached to C-3 (11) . Another example of a saponin present in plants as an inactive precursor is hederosaponin C of Hedera helix (7) .

Thus, some saponins exist as bi- and tridesmodic derivatives and are activated by conversion to the corresponding monodesmosidic form. Other saponins such as avenacin A-l and o;-tomatine (Figure 3) exist naturally as monodesmosidic saponins. It has been demonstrated that the presence of the sugar moiety at the C-3 carbon atom of the saponin molecule is important for biological activity, e.g. for the ability to complex with sterols and interfere with membrane integrity (1) , for toxicity and antimicrobial activity (2,7,9) and in the case of cucurbit saponins, for bitterness in taste (9) . Because many saponins have pronounced antifungal properties, they have been implicated in determining the resistance of plants to attack by saponin-sensitive fungi (8) and are regarded as being likely to be the most important class of preformed antifungal substances found in plants. While some plant pathogenic fungi have intrinsic resistance to the membranolytic action of saponins because of their membrane composition, others produce specific saponin-detoxifying enzymes, which are saponin glycosyl hydrolases (8) . These enzymes hydrolyse monosaccharides or oligosaccharides from the sugar chain at the C-3 carbon position of the saponin to give either a deglycosylated intermediate or the aglycone, often with consequent reduction in the toxicity of the saponin. Plants have also been reported to produce other saponin glycosyl hydrolases. For example tomato fruits have a tomatine glycosyl hydrolase activity which is involved in the breakdown of tomatine during fruit ripening (12) .

Saponin glycosyl hydrolases from fungi and plants described prior to the start of our work are listed in Table 2. These saponin glycosyl hydrolases are specific for the C-3 linked sugar moiety. It has been postulated that fungal saponin glycosyl hydrolases may be important in order for plant pathogenic fungi to overcome host defence mechanisms based on saponins (8, 15,-19) . Their mechanisms for detoxifying saponins may involve removal of one or more sugars. These enzymes have not been purified, and the genes encoding these enzymes have not been characterised (8,12,15-19,23-25) . Their possible roles in overcoming plant resistance have not been tested genetically.

Glycosyl hydrolases are known to have activities towards a range of carbohydrate-containing molecules. General classification of these has put the various enzymes into families on the basis of similarities at the nucleotide and/or amino acid level, but these enzymes have not been reported to have saponin (glycosyl) hydrolase activity (13, 14) .

As is apparent, hydrolase enzymes which can act on saponin molecules to reduce or remove their activity are of great utility and industrial applicability. The present invention provides for the first time, purified saponin glycosyl hydrolases and encoding nucleic acid therefor, and novel methods for screening nucleic acid, eg cDNA or genomic DNA, libraries of organisms for the presence of sequences encoding saponin glycosyl hydrolases, and cloning genes encoding such enzymes. The present invention is founded to a large extent on the unexpected discovery of similarities between different saponin glycosyl hydrolases from different organisms in terms of nucleotide sequences of the encoding genes, the amino acid sequences of the encoded enzymes, or the physicochemical properties or serological relatedness of the enzymes, notwithstanding differences previously observed or expected. The relatedness of enzymes from different organisms allows essentially the same protein purification procedure to be used for each enzyme. Previously, none of the enzymes have been purified sufficiently for amino acid sequencing. Furthermore, results described herein also provide the first direct evidence that saponin glycosyl hydrolases from plant pathogenic fungi overcome saponin- based antifungal defence mechanisms in plants and consequently determine host range of fungal pathogens. (See Bowyer et al . , 1995, Science 267, 371-373) .

As described in further detail below, in the research leading to the present invention, two saponin glycosyl hydrolase enzymes were purified to homogeneity from plant pathogenic fungi. These were (A) avenacinase, from the pathogen or cereal roots, Gaeumannomyces graminis var. avenae (Gga) , and (B) tomatinase, from the foliar pathogen of tomato, Septoria lycopersici (S. lycopersici) . Avenacinase detoxifies the triterpenoid saponin avenacin A-l

(Figure 3) found in oat roots, by hydrolytic cleavage of the β , 1-2 and /3,l-4 linked terminal glucose molecules to give sequentially the mono- and bis-deglucosylated forms, both of which are substantially less toxic to fungal growth than avenacin A-l itself. The action of the tomatinase enzyme from S. lycopersici is mechanistically similar to that of avenacinase, and involves hydrolytic cleavage of the terminal β,l-2 linked glucose from α-tomatine (Figure 3) . This deglycosylation is sufficient to destroy the ability of the saponin to complex with membrane sterols. Surprisingly, these two enzymes from diverse fungi and with clearly different substrate specificities were found to have very similar physicochemical properties with regard to fractionation using a range of protein separation techniques, identical pis, very similar molecular weights as determined by SDS-PAGE, and both were recognised by anti-avenacinase antiserum (Table 3) . It was a surprising discovery that antiserum raised against a specific saponin glycosyl hydrolase recognised a saponin glycosyl hydrolase from a taxonomically remote fungus. Another unexpected finding was that saponin glycosyl hydrolases in other organisms could also be identified and discriminated from other glucosidases by screening cDNA or genomic DNA libraries from such organisms with a suitable saponin glycosyl hydrolase cDNA probe. A third discovery is that saponin glycosyl hydrolases, although quite different in their substrate specificity, are highly homologous in nucleotide and amino acid sequence and in their physicochemical and serological properties .

The nucleotide and predicted amino acid sequences of the saponin glycosyl hydrolases avenacinase and tomatinase have been found to be closely related to jδ-glucosidases (EC 3.2.1.21) from the family 3 group identified by Henrissat in his classification of glycosyl hydrolases (13, 14) . The most closely related members of this group were BGL1 from Trichoderma reesei, BGL1 and BGL2 from Saccharomycopsis fibuligera (20) and BGLS from Candida pelliculosa (Figures 5 and 6) . It is stressed that activity of the members of the family 3 group of /3-glucosidases towards saponins has not been described. The family 3 /3-glucosidases are clearly distinct from other β-glucosidases described in Henrissat' s classification of glycosyl hydrolases (13, 14) on the basis of DNA sequence comparisons.

According to the present invention there is provided nucleic acid comprising a sequence encoding a saponin glycosyl hydrolase. The sequence may encode avenacinase, tomatinase or any other protein with saponin glycosyl hydrolase activity. Other closely related enzymes with similar physicochemical properties and which cross-react with anti-avenacinase antiserum have been identified in other varieties of

Gaeumannomyces graminis, and have been characterised in detail from G. graminis varieties tri tici and graminis . These enzymes weakly deglucosylate avenacin, and are referred to as avenacinase-like proteins (ALPs) . The genes encoding these ALPs are also homologous to the sequences encoding avenacinase from G. graminis var. avenae and tomatinase from S. lycopersici . The sequence may encode a polypeptide comprising an amino acid sequence encoded by any encoding sequence shown in Figure 4 and Figure 5 excluding those which are not saponin glycosyl hydrolases. Preferred amino acid sequences are shown in Figures 4 and 6. Preferably a sequence encoding avenacinase comprises all or part of the sequence shown in Figure 4c. Preferably, a sequence encoding tomatinase comprises all or part of the sequence shown in Figure 4b. The nucleic acid may comprise a sequence which is a variant, mutant, derivative or allele of a sequence which occurs naturally ("wild-type"), having a difference or alteration which is one or more of insertion, deletion and substitution of one or more nucleotides. The difference or alteration may or may not result in a change in amino acid sequence of the encoded protein. If there is a change in amino acid sequence of the encoded protein, the protein preferably retains saponin glycosyl hydrolase activity. Preferably the nucleic acid comprises a sequence with a high degree of homology to any of the sequences shown. High homology may be indicated by ability of complementary nucleic acid to hybridise under appropriate conditions for instance conditions stringent enough to exclude hybridisation to sequences not encoding a saponin glycosyl hydrolase.

The nucleic acid may be DNA or RNA and may be synthetic, eg with optimised codon usage for expression in a host organism of choice. The term "nucleic acid isolate" includes synthetic nucleic acid. DNA may be cDNA or genomic DNA.

Anti-sense nucleic acid is also provided by the invention, for example for use in gene regulation.

Thus, for instance, (anti-sense) nucleic acid according to the present invention may comprise a sequence of nucleotides complementary to a nucleotide sequence encoding an amino acid sequence shown in Figure 4, or an allele, variant, derivative or mutant thereof (as discussed) . Nucleic acid may be under the control of regulatory sequences for expression of anti-sense RNA for control of expression of a polypeptide such as a polypeptide with saponin glycosyl hydrolase activity. Host cells comprising nucleic acid for expression of anti-sense RNA and/or anti-sense RNA are provided by the invention, as are methods of production of anti- sense RNA comprising expression from a suitable DNA construct. Also provided by the present invention are a vector comprising nucleic acid as set out above, particularly any expression vector comprising appropriate regulatory sequences and from which the encoded polypeptide can be expressed under appropriate conditions, a host cell comprising or containing (e.g.transformed) with any such vector (such as a plant 'or microbial cell, including fungal) and an organism (e.g. fungus or plant) containing such a host cell. The work described herein provides the first purification of any saponin glycosyl hydrolase. Thus, the present invention provides an isolated saponin glycosyl hydrolase, i.e. one in substantially pure form or free from contaminants with which it is naturally associated. The term "purified to homogeneity" may be used.

Embodiments of this include avenacinase and tomatinase and others comprising all or part of any of the sequences shown in the figures. Also provided are polypeptides which comprise mutants, alleles, variants and derivatives of polypeptides with any of the sequences shown in the figures, as a result of addition, insertion, deletion or substitution of one or more amino acids. In purification/isolation methods according to the present invention, a crude extraction with possible subsequent concentration, eg by precipitation, is followed by iso-electric focusing. The iso-electric focusing may be followed by anionic exchange and/or gel filtration.

A convenient way of producing a saponin glycosyl hydrolase, according to the present invention, is to express nucleic acid encoding it . Accordingly, the present invention also encompasses a method of making a polypeptide which has saponin glycosyl hydrolase activity, the method comprising expression of nucleic acid encoding the polypeptide. This may conveniently be achieved by growing a host cell, transfected with a vector comprising such nucleic acid, under conditions which cause or allow expression of the polypeptide. Alternatively expression may be in an in vi tro system. Systems for cloning and expression of a polypeptide in a variety of different host cells are well known. Suitable host cells include bacteria, mammalian cells, yeast, fungi, plant and baculovirus systems. One preferred host is Neurospora crassa . Suitable vectors can be chosen or constructed, containing appropriate regulatory sequences, including promoter sequences, terminator fragments, polyadenylation sequences, enhancer sequences, marker genes and other sequences as appropriate. For further details see, for example, Molecular Cloning: a

Laboratory Manual : 2nd edition, Sambrook et al, 1989, Cold Spring Harbor Laboratory Press. See eg Jones et al (26) for review of vectors for transformation and expression in plants. A suitable vector for expression in Neurospora crassa is pBLUESCRIPT II KS (Stratagene) . An alternative vector is pIAT, which is an expression vector consisting of the N. crassa isocitrate lyase gene promoter, a multiple cloning site, and the N. crassa glutamate dehydrogenase gene terminator in pBLUESCRIPT (Figure 7) .

The invention further encompasses a host cell transformed with or comprising a vector which comprises nucleic acid encoding a polypeptide with saponin glycosyl hydrolase activity, especially a plant, fungal or microbial cell. Thus, a host cell, such as a plant, fungal or microbial cell, comprising nucleic acid according to the present invention is provided. Within the cell, the nucleic acid is preferably incorporated within the chromosome.

A vector comprising nucleic acid according to the present invention need not include a promoter, particularly if the vector is to be used to introduce the nucleic acid into cells for recombination into the genome.

The present invention further encompasses a plant comprising a plant cell comprising nucleic acid according to the present invention, seed, selfed or hybrid progeny and any clone or descendant of such a plant, also any part of such a plant, progeny, clone or descendant, e.g. cuttings, seed. The invention provides any plant propagule, that is any part which may be used in reproduction or propagation, sexual or asexual, including cuttings, seeds and so on.

Also according to the invention there is provided a plant cell having incorporated into its genome a sequence of nucleotides as provided by the present invention, preferably under operative control of a promoter for control of expression of the encoded polypeptide. A further aspect of the present invention provides a method of making such a plant cell involving introduction of a vector comprising the sequence of nucleotides into a plant cell and causing or allowing, under suitable conditions, recombination between the vector and the plant cell genome to introduce the sequence of nucleotides into the genome.

Transformation procedures depend on the host used, but are well known.

When introducing nucleic acid into a cell, certain considerations must be taken into account, well known to those skilled in the art. There rfiust be available a method of transporting the construct into the cell. Once the construct is within the cell membrane, integration into the endogenous chromosomal material either will or will not occur. Plants transformed with the DNA segment containing pre-sequence may be produced by standard techniques which are already known for the genetic manipulation of plants. DNA can be transformed into plant cells using any suitable technology, such as a disarmed Ti-plasmid vector carried by Agrobacterium exploiting its natural gene transfer ability (EP-A-270355, EP-A-0116718, NAR 12(22) 8711 - 87215 1984) , particle or microprojectile bombardment (US 5100792, EP-A-444882, EP-A-434616) microinjection (WO 92/09696, WO 94/00583, EP 331083, EP 175966) , electroporation (EP 290395, WO 8706614) or other forms of direct DNA uptake (DE 4005152, WO 9012096, US 4684611) . Agrobacterium transformation is widely used by those skilled in the art to transform dicotyledonous species. Although Agrobacterium has been reported to be able to transform foreign DNA into some monocotyledonous species (WO 92/14828) , microprojectile bombardment, electroporation and direct DNA uptake are preferred where Agrobacterium is inefficient or ineffective. Alternatively, a combination of different techniques may be employed to enhance the efficiency of the transformation process, eg bombardment with Agrobacterium coated microparticles (EP-A-486234) or microprojectile bombardment to induce wounding followed by co-cultivation with Agrobacterium (EP-A-486233) .

The particular choice of a transformation technology will be determined by its efficiency to transform certain plant species as well as the experience and preference of the person practising the invention with a particular methodology of choice. It will be apparent to the skilled person that the particular choice of a transformation system to introduce nucleic acid into plant cells is not essential to or a limitation of the invention.

The proteins provided by the present invention may be purified from natural sources, or be produced recombinantly. Such purified proteins and methods of their purification, from natural sources or recombinantly produced, are encompassed by the present invention. Indeed the present invention extends to any of the proteins obtained or obtainable using a method according to the present invention.

Polypeptides with saponin glycosyl hydrolase activity may, in accordance with the invention, be used in deglycosylating a saponin in a method which comprises contacting the saponin with the polypeptide. The contacting may be in vivo or ex vivo ( in vi tro) . Contact may in particular be following earlier expression of the hydrolase enzyme e.g. by cells of a plant or fungus ( in vivo) . Purified protein may be used to raise antibodies employing techniques which are standard in the art. Methods of producing antibodies include immunising a mammal (eg human, mouse, rat, rabbit, horse, goat, sheep or monkey) with the protein or a fragment thereof. Antibodies may be obtained from immunised animals using any of a variety of techniques known in the art, and might be screened, preferably using binding of antibody to antigen of interest. For instance, Western blotting techniques or immunoprecipitation may be used (Armitage et al, 1992, Nature 357: 80-82) .

It has been found surprisingly that antibodies raised to a saponin glycosyl hydrolase can be used in the identification and/or isolation of other saponin glycosyl hydrolases. Thus, the present invention provides a method of identifying or isolating a polypeptide with saponin glycosyl hydrolase activity, comprising screening candidate polypeptides with antibody which is able to bind a saponin glycosyl hydrolase, or preferably has binding specificity for a saponin glycosyl hydrolase. The saponin glycosyl hydrolase may be any of those with sequences shown in Figure 4.

Candidate polypeptides for screening may be the products of an expression library, eg a lambda library, created using nucleic acid derived from an organism of interest, such as one expected to produce a saponin glycosyl hydrolase, or may be the product of a purification process from a natural source.

A polypeptide found to bind an antibody specific for a saponin glycosyl hydrolase, such as avenacinase, may be isolated and then may be subject to amino acid sequencing. Any suitable technique may be used to sequence the polypeptide either wholly or partially (for instance a fragment of the polypeptide may be sequenced) . Amino acid sequence information may be used in obtaining nucleic acid encoding the polypeptide, for instance by designing one or more oligonucleotides (e.g. a degenerate pool of oligonucleotides) for use as probes or primers in hybridisation to candidate nucleic acid, as discussed further below.

The antibody may be raised by immunisation, as discussed above. In principle it may be polyclonal or monoclonal. In one embodiment of the present invention, polyclonal antisera raised against avenacinase is used to identify another saponin glycosyl hydrolase.

As an alternative or supplement to immunising a mammal, antibodies with appropriate binding specificty may be obtained from a recombinantly produced library of expressed immunoglobulin variable domains, eg using lambda bacteriophage or filamentous bacteriophage which display functional immunoglobulin binding domains on their surfaces; for instance see WO92/01047. The library may be naive, that is constructed from sequences obtained from an organism which has not been immunised with any protein of interest (or fragments) , or may be one constructed using sequences obtained from an organism which has been exposed to the antigen of interest.

It is possible to take monoclonal antibodies and use the techniques of recombinant DN_ technology to produce other antibodies or chimeric molecules which retain the specificity of the original antibody. Such techniques may involve introducing DNA encoding the immunoglobulin variable region, or the complementarity determining regions (CDRs) , of an antibody to the constant regions, or constant regions plus framework regions, of a different immunoglobulin. See, for instance, EP-A_^184187, GB 2188638A or EP-A-239400. A hybridoma producing a monoclonal antibody may be subject to genetic mutation or other changes, which may or may not alter the binding specificity of antibodies produced.

As antibodies can be modified in a number of ways, the term "antibody" should be construed as covering any specific binding substance having an binding domain with the required specificity. Thus, this term covers antibody fragments, derivatives, functional equivalents and homologues of antibodies, including any polypeptide comprising an immunoglobulin binding domain, whether natural or synthetic. Chimaeric molecules comprising an immunoglobulin binding domain, or equivalent, fused to another polypeptide are therefore included. Cloning and expression of chimaeric antibodies are described in EP-A-0120694 and EP-A-0125023.

It has been shown that fragments of a whole antibody can perform the function of binding antigens. Examples of binding fragments are (i) the Fab fragment consisting of VL, VH, CL and CHI domains; (ii) the Fd fragment consisting of the VH and CHI domains; (iii) the Fv fragment consisting of the VL and VH domains of a single antibody; (iv) the dAb fragment (Ward, E.S. et al . , Nature 341, 544-546 (1989)) which consists of a VH domain; (v) isolated CDR regions; (vi) F(ab')2 fragments, a bivalent fragment comprising two linked Fab fragments (vii) single chain Fv molecules (scFv) , wherein a VH domain and a VL domain are linked by a peptide linker which allows the two domains to associate to form an antigen binding site (Bird et al, Science, 242, 423-426, 1988; Huston et al, PNAS USA,

85, 5879-5883, 1988) ; (viii) bispecific single chain Fv dimers (PCT/US92/09965) and (ix) "diabodies" , multivalent or multispecific fragments constructed by gene fusion (WO94/13804; P. Holliger et al Proc. Natl. Acad. Sci. USA 90 6444-6448, 1993) .

The antibody may be labelled using any one of a number of techniques known in the art.

The sequences of nucleic acid encoding different saponin glycosyl hydrolases have been found to be sufficiently conserved to enable nucleic acid encoding one such enzyme (eg from one organism) to be used as a probe to identify and/or isolate nucleic acid encoding another such enzyme (eg from a different organism) . This success is despite the very different substrate specificities observed when saponin glycosyl hydrolases of different organisms have been compared.

Thus, the present invention also provides a method of identifying and/or isolating nucleic acid encoding a polypeptide with saponin glycosyl hydrolase activity, comprising probing candidate (or "target") nucleic acid with nucleic acid which encodes a saponin glycosyl hydrolase or encodes a fragment thereof. The candidate nucleic acid (which may be, for instance, cDNA or genomic DNA) may be derived from any cell or organism which may contain or is suspected of containing nucleic acid encoding a saponin glycosyl hydrolase. Example sources are listed in Table 2. Furthermore, the method enables the first identification of the existence of a saponin glycosyl hydrolase in an organism and perhaps also the identification of a new saponin or a saponin acting in a manner not previously known.

For instance, it may be that an organism is pathogenic for a particular plant but that another organism is not. This might be because of the presence of a saponin in the plant. Rather than trying to determine the presence of a saponin, probing the nucleic acid of the pathogenic organism in a manner according to the present invention may lead to the identification of a gene encoding a saponin glycosyl hydrolase within the genome of the organism. The absence of an equivalent gene within the genome of the other, non-pathogenic organism would point towards the pathogenicity being as a result of the hydrolase being used to overcome the defence mechanism of the plant employing a saponin.

In a preferred embodiment of the present invention, the nucleic acid used for probing of candidate nucleic acid encodes avenacinase, tomatinase, any of the amino acid sequences shown in Figure 4, a sequence complementary to a coding sequence, or a fragment of any of these. Preferred nucleotide sequences appear in Figures 4 and 5. Fragments of any of these may be used.

Preferred conditions for probing are those which are stringent enough for there to be a simple pattern with a small number of hybridisations identified as positive which can be investigated further. It is well known in the art to increase stringency of hybridisation gradually until only a few positive clones remain. This aspect of the present invention is based on the surprising discovery of the relatedness of different saponin glycosyl hydrolases in terms of homology of the sequences of encoding nucleic acid. Furthermore, it has been found that when genomic libraries of organisms are probed using the nucleic acid of one saponin glycosyl hydrolase, only a few clones are identified as potentially encoding a saponin glycosyl hydrolase. This, too, is surprising given the close relationship between saponin glycosyl hydrolases and -glucosidases, which are abundant.

Probing according to the present invention enables discrimination of enzymes which have saponin glycosyl hydrolase activity from other β-glucosidases which do not have this activity. Thus, /3-glucosidases previously not known to have saponin glycosyl hydrolase activity may be identified as having this activity as a result of hybridisation of encoding DNA with a probe according to the present invention. As an alternative to probing, though still employing nucleic acid hybridisation, oligonucleotides designed to amplify DNA sequences from genes encoding saponin glycosyl hydrolases (but not from family 3 3-glucosidases with no saponin-degrading activity) may be used' in PCR reactions or other methods involving amplification of nucleic acid, using routine procedures. See for instance "PCR protocols; A Guide to Methods and Applications", Eds. Innis et al, 1980, Academic Press, New York.

Preferred amino acid sequences suitable for use in the design of probes or PCR primers are sequences conserved (completely, substantially or partly) between saponin glycosyl hydrolases, for example two or more, preferably selected from avenacinase, tomatinase and the ALP's disclosed herein. Preferred positions are indicated in Figure 6 (amino acids) and Figure 5 (nucleotides) by means of arrows.

On the basis of amino acid sequence information, oligonucleotide probes or primers may be designed, taking into account the degeneracy of the genetic code, and, where appropriate, codon usage of the organism from the candidate nucleic acid is derived. Preferred nucleotide sequences may include those comprising or having a sequence encoding amino acids A T H S/A G V (see Figure 6) or complementary to a nucleotide sequence encoding amino acids D P D/Y V/A W Q A D (see Figure 6) . Preferred oligonucleotide sequences include G G G/T A/C C G C C/T A C C C A T/C T/G C A/C G G A and T C G G C T/C T G C C A A/C A/G C A/G T A/C T/C G G G T C C (both sequences given 5' to 3' ; see Figure 5) . Preferably an oligonucleotide in accordance with the invention, e.g. for use in nucleic acid amplification, has about 10 or fewer codons (e.g. 6, 7 or 8) , i.e. is about 30 or fewer nucleotides in length (e.g. 18, 21 or 24) . Thus, in accordance with the present invention, an enzyme may be identified as a saponin glycosyl hydrolase without the need for there to be a saponin substrate known for that enzyme.

The invention enables the identification of specific saponin glycosyl hydrolases of important pathogens and subsequent use of such enzymes to screen for and/or rationally design synthetic ligands, e.g. chemicals, peptides (which may be synthetic) , capable of inhibiting the enzyme activity, thereby reducing the pathogenicity of such pathogens for specific hosts.

Synthetic peptides which bind to tomatinase have been identified by screening a random hexapeptide phage display library (Scott and Smith, Science, 249, 389-390, 1990) . A number of recombinant phage were isolated by virtue of their ability to bind tomatinase. Some of these phage also inhibited tomatinase enzyme activity (see example 4 and Figure 8) . Such enzyme inhibiting ligands may be used as an active ingredient in agro-chemical formulations that can be applied to control plant pathogens.

Accordingly, the invention further provides a method of screening for a molecule able to modulate saponin glycosyl hydrolase activity, comprising contacting a candidate molecule or candidate molecules with a polypeptide as provided herein, and determining activity. Modulation may increase (augment) or decrease (inhibit) saponin glycosyl hydrolase activity, of one or more saponin glycosyl hydrolases. The molecules may be peptides or polypeptides, for instance, and may be screened using any suitable technique, such as bacteriophage display. An initial screening step may be determination of binding of a candidate molecule to a saponin glycosyl hydrolase.

The present invention also provides molecules able to modulate activity of a saponin glycosyl hydrolase ("modulators", which may be activators/enhancers or inhibitors) . These may be obtainable using a screening method as disclosed herein. For example, a modulator of saponin glycosyl hydrolase activity may be identified using a screening method as set out in the preceding paragraph.

Modulator molecules which are peptides or polypeptides may be expressed from encoding nucleic acid therefor, in vi tro or in vivo using standard techniques. Expression may take place in a host cell, e.g. a plant or microbial (e.g. fungal) cell, and this may follow earlier introduction of the encoding nucleic acid into the cell . A plant host cell may be within a plant. Thus, plants comprising such a host cell, clones, descendants and any part thereof, including any propagule as discussed above, are encompassed by the present invention.

On the other hand, molecules, including peptides and polypeptides, may be synthetic, e.g. made using any standard synthesis technique. Further provided by the present invention is a method of modulating activity of a saponin glycosyl hydrolase, comprising contacting the saponin glycosyl hydrolase with a molecule which is a modulator of activity of a saponin glycosyl hydrolase. This may follow expression of the modulator from encoding nucleic acid therefor, as disclosed. It may follow identification of the molecule as being able to modulate saponin glycosyl activity using a method in accordance with the present invention. Thus, an aspect of the present invention provides the use of a molecule in modulation of saponin glycosyl hydrolase activity of a polypeptide, as disclosed. The molecule may be selected from the peptides shown in Figure 8, or may comprise a peptide selected from those shown in Figure 8. Of course, the amino acid sequences of the peptides may be subject to alteration by any of addition, insertion, deletion and substitution of one or more amino acids, and any molecule comprising such an altered amino acid sequence is encompassed by the present invention, as long as it is able to modulate activity of a saponin glycosyl hydrolase.

The invention enables modification of the host range specifity of plant microbes . This could be exploited, for example, to engineer suitable microbial organisms to be capable of colonising hosts who will benefit from such an interaction. See eg refs 27-32. The invention also allows the identification and production of enzymes -or enzyme extracts with a specific saponin glycosyl hydrolase activity for direct use in industrial processes . A slightly different application may be the engineering of micro-organisms that are used in industrial processes and that excrete saponin glycosyl hydrolysing enzymes of interest. For example, saponin glycosyl hydrolases could be cloned from plant pathogenic microbes that are capable of hydrolysing specific saponins in plants used for food or feed stuffs to modify the taste by reducing bitterness. Examples of such applications include hydrolysing soybean saponins that have bitter and astringent tastes (refs. 21,22) during the production of soy sauce, or hydrolysing the problematic bitter saponins in the processing of sugarbeets and the legumes (1) . Alternatively, the enzymes can be used to reduce the anti-nutritional composition or toxicity of specific crops, such as, for example, glyco-alkaloids in Solanaceous species or digitoxin in Foxglove. Another useful application of the enzymes may be the development of biochemical assay kits or devices such as bio-sensors for measuring specific saponin concentrations. There is also considerable interest in the use of saponin deglycosylating enzymes as chemical agents to study the structures of saponins or the degradation thereof to provide for new materials for chemical synthesis of pharmaceuticals etc. (refs. 6,7,9) . Food or feed crops may be modified by expressing a suitable saponin glycosyl hydrolase under the control of a tissue or developmental specific promoter to modify the saponin contents of crops in si tu or alternatively in a process dependent manner. This can be brought about by targeting the enzymes to a cellular compartment where they are combined with respectively separated from the substrate by methods well described in the literature. Processing of crops where enzyme and substrate are engineered to be physically separated would result in the breakdown of such barriers resulting in a desired catalytic reaction. Such crops could have enhanced value as a food or feed crop, or optionally by accumulating different intermediates or end products provided new compounds for the speciality chemical and pharmaceutical industry. Alternatively, expression of genes coding for a saponin glycosyl hydrolase homologue to suppress endogenous plant enzyme activity resulting in the accumulation of intermediates in the saponin biosynthetic pathways may provide new materials for industry.

The invention also provides a method to identify and clone the genes encoding saponin glycosyl hydrolase from specific plants and subsequently introduce genetic constructs into such plants to regulate down or up the expression of endogenous saponin glycosyl hydrolase activity thereby altering the nutritional composition or increasing the concentration of certain desirable intermediates or end products in the biosynthesis of saponins. For example, ginseng, ginger and liquorice are full of saponins. Increasing or decreasing, the saponin concentration or altering the saponin content qualitatively, either through the genetic suppression of plant saponin glycosyl hydrolases or through selecting for progeny with increased or reduced enzyme activity in a breeding strategy may be beneficial.

The invention also provides methods to develop mutant plant pathogens with reduced pathogenicity which could be used to screen for plants in breeding programs that produce saponins or higher levels of saponins.

Illustrative embodiments of the present invention are discussed further below, with reference to the Figures.

Figure 1 shows four aglycones on which saponin structures are based: 1(a) - triterpene; 1(b) - spirostanol; 1(c) - alkaloid; 1(d) - furostanol.

Figure 2 shows avenacosides A and B, which are inactive bisdesmosidic saponins found in intact oat leaves. Figure 3: (a) shows the structure of avenacin A-l; (b) shows the structure of a-tomatine.

Figure 4 shows DNA and predicted amino acid sequences of tomatinase from Septoria lycopersici , avenacinase from G . graminis var. avenae, and the avenacinase-like proteins (ALPs) from G. graminis vars. tri tici and graminis : Figure 4(a) shows a Septoria lycopersici tomatinase cDNA sequence; Figure 4(b) shows a Septoria lycopersici tomatinase amino acid sequence derived from cDNA; Figure 4(c) shows a G. graminis var. avenae avenacinase genomic DNA sequence and predicted amino acid sequence; Figure 4(d) shows a G. graminis var. graminis avenacinase-like protein (ALP) genomic DNA sequence and predicted amino acid sequence; Figure 4(e) shows a G. graminis var. tri tici ALP genomic DNA sequence and predicted amino acid sequence.

Figure 5 shows alignment of the coding nucleotide sequence of tomatinase cDNA from Septoria lycopersici (Tom) with the predicted coding sequences of avenacinase from G. graminis var. avenae (avenacinase) , the avenacinase-like proteins (ALPs) from G. graminis vars. tritici and graminis, and other related /_ -glucosidases present in the computer databases, specifically members of the family 3 group of β- glucosidases : Trichoderm = BGL1 from Trichoderma reesei ; BGL1 and BGL2 are /_ -glucosidases from S ac cha omy cops is fibuligera; Cand - BGLS from Candida pelliculoεa ; (EMBL accession numbers Tr09580, Sfglua, Sfglub and Cpglucb, respectively) . Sequences were aligned using the PILEUP algorithm (Devereux et al. , 1984 Nuc . Acids Res . 12, 387-395) . Nucleotides found in the majority of sequences are highlighted in black. The positions of sequences which would be suitable for the generation of PCR primers are indicated by arrows. Figure 6 shows alignment of the predicted amino acid sequences of the DNA sequences presented in Figure 5. Identical amino acids are boxed in black. Conserved amino acids are shaded. Figure 7 shows the Neurospora crassa expression vector, pIAT. This vector contains a promoter from the N. crassa isocitrate lyase gene, a multiple cloning site, and a terminator from the N. crassa glutamine dehydrogenase gene, ligated into pBLUESCRIPT. Figure 8 shows inhibition of Septoria lycopersici tomatinase activity by peptides expressed on the surface of bacteriophage particles . Tomatinase activity in the presence of equimolar phage. Letters indicate the deduced amino acid sequence of the displayed hexapeptide as determined by DΝA sequencing. The bars represent tomatinase activity measured by the ability to deglucosylate methyl umbelliferyl glucoside. The error bar shows the predicted variation in unmodified tomatinase activity.

EXAMPLE 1

PURIFICATION OF SAPONIN GLYCOSYL HYDROLASES

The saponin glycosyl hydrolase from Gga, avenacinase, was purified to homogeneity. Problems were encountered in purifying avenacinase due to the low abundance of the protein and its susceptibility to degradation by proteolytic enzymes during fractionation. The following procedure was finally adopted, which differs from the preliminary method for partial characterisation described in Osbourn et al (1991) (Physiol. Mol. Plant. Pathol. 38, 301-312) .

Proteins were precipitated from 40 1 of culture filtrate with ammonium sulphate after addition of phenylmethylsulphonyl fluoride and EDTA (final concentrations 0.05 mM and 2 mM respectively) . Dialysed protein preparations were made up to a volume of 100 ml containing 10% glycerol, 0.5 % Nonidet P-40 and 1% ampholytes, pH range 4-6.5 (Pharmalyte) , and separated by a recycling free flow IEF using the RF3

Protein Fractionator (Texas Instruments) . Avenacinase activity was assayed by monitoring avenacin deglycosylation using thin layer chromatography according to Osbourne et al. Active fractions were pooled, de-salted and exchanged into 20 mM sodium phosphate buffer pH 7.0, 0.2 M sodium chloride using Centricon C-30 spin columns prior to gel filtration on a 7.8 x 250 mm TSK G3000 SW XL HPLC column in the same buffer. The homogeneity of the final purified avenacinase preparation was confirmed by SDS-PAGE. The avenacinase-like protein (ALP) from G. graminis var. tri tici was purified in exactly the same way, and was assayed for activity towards avenacin, which is weak.

Tomatinase was purified from culture filtrates of S . lycopersici using essentially the same purification procedure. Proteins from filtrates of S. lycopersici cultures were concentrated with ammonium sulphate and dialysed as follows.

Liquid cultures of S. lycopersici were filtered through Miracloth (Calbiochem, La Jolla, USA) . The following protease inhibitors were added to the filtrate to the indicated concentrations: phenylmethylsulphonyl fluoride (0.05 mM) , EDTA (2 mM) , benzamidine hydrochloride (1 μM) , phenanthroline (0.5 μM) , aprotinin (0.5 μM) , leupeptin (2 μM) and pepstatin A (1.5 μM) . The preparation was chilled to 4°C, and after 1 h proteins were precipitated by the addition of ammonium sulphate to 80% saturation as described in Osbourn et al . (1991) .

The dialysate was then fractionated by free-flow isoelectric focussing, and a single peak of tomatinase activity with a pi of 4.6 was identified. Active fractions were pooled and subjected to high-performance anion exchange chromatography on DEA-SPW at pH 6.2; all the tomatinase activity bound to the column. A gradient of 0 to IM sodium chloride was applied to the column and tomatinase was eluted as a single peak at a salt concentration of 0.15 M sodium chloride. Fractions containing tomatinase activity were then pooled and further fractionated by high performance size exclusion chromatography. Again, tomatinase activity eluted as a single peak. SDS-PAGE of these fractions indicated that the enzyme had been purified to homogeneity and was a large protein of molecular weight approximately 113 kD (Fig. la) .

The tomatinase enzyme activity was assayed by measurement of glucose release from α-tomatine. Samples were assayed using 250 μm or 2mM -tomatine (Sigma) in 100 mM sodium acetate buffer pH 5 at 37°C for 15 min to 2h. Reactions were stopped by boiling for 10 min, and glucose release measured using the glucose oxidase assay (Sigma) . Tomatinase was found to fractionate in the same way as avenacinase when the techniques of isoelectric focusing, anion exchange and size exclusion HPLC and SDS-PAGE were employed. For amino acid sequence analysis, tomatinase protein was cleaved with the proteolytic enzymes Asp-N or Lys-C (Boehringer Mannheim) .

The physicochemical parameters of tomatinase assessed during this purification procedure are very similar to those of avenacinase. The two enzymes have identical pis (4.6) and very similar molecular weights (110 kD and 113 kD for avenacinase and tomatinase respectively) .

Furthermore, the purified protein was immunologically related to avenacinase. Polyclonal antisera raised against avenacinase from G. graminis var. avenae (see Example 2) recognised purified tomatinase, and a single protein species of the same molecular weight as tomatinase in crude protein preparations from culture filtrates of S. lycopersici . Estimation of the relative amounts of tomatinase activity and of the degree of immunological reactivity in samples from different stages of the purification were consistent with tomatinase being the only immunologically reactive protein present in the original culture filtrate. There was no evidence to indicate the existence of other S. lycopersici proteins which were immunologically cross-reactive with avenacinase.

EXAMPLE 2

CLONING, EXPRESSION AND MANIPULATION OF A SAPONIN GLYCOSYL HYDROLASE GENE

The gene encoding avenacinase was cloned by raising rat polyclonal antisera to the purified enzyme and using this to screen a lambda gtll cDNA expression library of Gga. cDNA clones were isolated and found to contain common restriction fragments, indicating that they were related. A genomic clone was obtained from Gg. var. avenae genomic DNA as follows. Genomic DNA was digested with a number of different restriction enzymes and the DNA fragments separated by gel electrophoresis. After Southern blot analysis using avenacinase cDNA as a probe, a 10.6 kb Bgl II fragment was identified for cloning purposes. A size-enriched library of Bgl II DNA fragments was made in the pUC derivative Bluescript (Stratagene) , and positive clones were identified by colony hybridisation with the avenacinase cDNA probe. The genomic clone and total genomic DNA from Gg. var. avenae were digested with a number of restriction enzymes. Hybridisation with a cDNA clone indicated the presence of a single copy of the avenacinase gene in the genome. The genomic clone was introduced into Neurospora crassa (which is avenacin A-l sensitive and avenacinase-minus) by DNA transformation, and demonstrated to confer resistance to avenacin and avenacinase activity. Gene disruption experiments were then carried out in which a plasmid bearing an internal portion of the gene and a selectable marker was transformed into Gga, and a number of avenacin-sensitive mutants were obtained. Southern blot analysis confirmed that the gene had been disrupted. The avenacin-sensitive mutants lacked the avenacinase protein as determined by enzyme assays and Western blot analysis, and showed clearly reduced pathogenicity to oats but retained wild type pathogenicity to wheat.

These experiments indicate a role for avenacinase in determining specificity of Gga to oats.

(a) Preparation of anti-avenacinase antisera Avenacinase preparations were made from filtrates of cultures of Gg var. avenae essentially as described by Osbourn et al . Protein concentrations were measured using the Biorad Protein Assay. Proteins were precipitated from 40 1 of culture filtrate with ammonium sulphate after addition of phenylmethylsulphonyl fluoride and EDTA (final concentrations 0.05 mM and 2 mM respectively) . After dialysing, the protein preparation was made up to a volume of 100 ml containing 10% glycerol, 0.5% Nonidet P-40 and 1% ampholytes, pH range 4-6.5 (Pharmalyte) , and separated by recycling free flow IEF using the RF3 Protein Fractionator (Texas Instruments) . Active fractions were pooled, de-salted and exchanged into 20 mM sodium phosphate buffer pH 7.0, 0.2 M sodium chloride using Centricon C-30 spin columns prior to gel filtration on a 7.8 x 250 mm TSK G3000 SW XL HPLC column in the same buffer. Active fractions were electroeluted from as SDS-PAGE gel and used to immunise 'Wistar' rats. The polyclonal antiserum was used for Western blot analysis and cDNA library screening at dilutions of 1:2000. Control experiments with pre-immune serum (diluted 1:1000) gave no signal. Rabbit anti-rat IgG alkaline phosphatase conjugate was used as the secondary antibody.

(b) cDNA library construction

RNA was prepared from mycelia grown in liquid cultures by the method of Chirgwin et al (1979 Biochemistry 18 64: 5294-5299) . Poly-A RNA was isolated from 2 mg of total RNA using oligo-dT spin columns. A cDNA expression library was constructed in lambda gtll and screened with the polyclonal antiserum. Positive plaques were isolated and the inserts amplified from plaques cored from plates. All cDNAs isolated in this way cross-hybridised on a Southern blot and were used as probes to isolate larger clones from the original cDNA library and from a genomic library constructed in pGM32. A 1.1 kb cDNA clone

(A312) and a 10.6 kb genomic clone (AG31) were selected for use in the subsequent experiments.

(c) Transformation of N. crassa

A pyr4- mutant of N. crassa was co-transformed with the plasmid pGM32 containing the pyr4 gene, and the plasmid PA3G1 containing the putative avenacinase genomic clone. Pyr4⁺ colonies were selected and tested for resistance to 10 mg/ml avenacin. Resistant colonies (4/30) were tested for avenacinase enzyme activity. DΝA was isolated from these resistant colonies and probed with A3G1 to demonstrate the presence of transforming DNA. Only colonies containing transforming A3G1 DNA showed avenacin resistance and avenacinase activity.

Southern blot analysis indicated that the transforming DNA had integrated into different positions in different transformants. Culture filtrates of these transformants had avenacinase activity and contained an immunologically cross-reactive protein of the same size as that found in Gg var. avenae culture filtrates. A single band of approximately 2.7 kb (consistent with the size of the avenacinase protein) was observed by Northern blot analysis of a number of Gg var. avenae isolates. DNA sequence analysis indicates clear relatedness to a class of fungal /5-glucosidases. (See Figures 5 and 6.)

(d) Mutation of Get var. avenae

Avenacin-sensitive mutants of Gg var. avenae were obtained by targeted disruption of the avenacinase gene. This technique relies on homologous recombination between transforming DNA and the fungal chromosome, resulting in an insertion of foreign DNA into the target gene. Such mutants were selected by Southern blot analysis and tested for their resistance to avenacin, production of the avenacinase protein, and for pathogenicity to oats and the non-saponin containing host, wheat. Transformants in which plasmid DNA had been integrated into the avenacinase gene did not have any detectable avenacinase activity, no longer produced protein which cross-reacted with the anti-avenacinase antisera, and were 4-5 times more sensitive to avenacin. Controls consisting of transformed fungi in which the avenacinase gene had remained intact had wild-type levels of avenacinase activity and retained the immunologically cross-reactive protein.

EXAMPLE 3 PROBING FOR SAPONIN GLYCOSYL HYDROLASE GENES AND ISOLATION THEREOF

Southern blot experiments using the avenacinase cDNA as a probe revealed that cross-hybridising DNA was present in genomic DNA of a number of fungi, including S lycopersici .

A full length cDNA clone encoding tomatinase from S lycopersici was obtained using avenacinase cDNA as a probe.

A cDNA library was made in the lambda ZAP II vector (Stratagene) from RNA preparations from S. lycopersici cultures producing tomatinase. This library was probed with a 1.1 kb avenacinase cDNA fragment. Six cDNA clones with insert DNA sizes of up to 2.5 kb were isolated. The S lycopersici DNA in these clones cross-hybridised and contained common restriction fragments, indicating that the cDNA clones were related. Southern blot analysis of S. lycopersici genomic DNA digested with a range of restriction enzymes and probed with the longest cDNA indicated only a single cognate gene (no other cross-hybridising DNA was detected in the S. lycopersici genome) . The shorter cDNA clones were assumed to be truncated forms and this was later confirmed by DNA sequence analysis of the ends of these clones.

The entire sequence of the tomatinase cDNA has been obtained (Figure 4) . Comparison of the amino acid sequence of the predicted product encoded by this cDNA with amino acid sequence information from tomatinase enzyme purified from S. lycopercici confirmed that the cDNA did indeed encode tomatinase (Figure 4.) .

This was further verified by the demonstration that expression of this cDNA in N. crassa conferred tomatinase activity and enhanced resistance to -tomatine.

A 2.5 kb Xhol restriction fragment extending from the X ol site at the beginning of the cDΝA to another site in the polylinker beyond the 3' end of the cDΝA was cloned into the N. crassa expression vector, pIAT (Figure 7) . This plasmid was tested for the ability to confer of-tomatine resistance and tomatinase activity when introduced into N. crassa , which is relatively sensitive to α-tomatine and has no detectable tomatinase activity. A uridine-requiring (pyr4 ) mutant of N. crassa was co-transformed with the plasmid pGM32 and the pIAT construct following the same procedure described in Example 2c. A proportion of the pyr4⁺ transformants were expected to have received the pIAT construct bearing the tomatinase cDNA.

Of 30 pyr4⁺ transformants recovered from this co-transformation, eight showed increased growth rates on α-tomatine when compared to transformants which had received pIAT without the cDNA insert. Four transformants grew at rates ranging from 1.7-2.6 times that of the controls, and formed colonies with regular even margins, while the growth rates of the remaining four were between 1.5 and 1.7 times higher than the controls, and the colonies were irregular in shape and had more aerial mycelium. This resistance to cϋ-tomatine was absolutely correlated with the presence of the pIAT construct containing the tomatinase cDNA

(as determined by Southern blot analysis) and with the presence of tomatinase activity.

The full DNA sequence of the avenacinase gene has been determined (Figure 4b) , and comparison of this sequence with that encoding tomatinase (Figure 4a) revealed approximately 60% nucleotide identity, with 53% identity and 68% similarity at the amino acid level (Figures 5 and 6, and Table 4) .

Two further enzymes belonging to this group (avenacinase-like proteins, or ALPs) have been characterised from the wheat and grass pathogens Gaeumannomyces graminis varieties tri tici and graminis . The ALP from G. graminis var. tri tici has been purified to homogeneity (Table 3) . The cognate genes from G. graminis var. tri tici and var. graminis have been isolated and the full nucleotide sequences determined (Figure 4d and e) . The G. graminis var. tri tici and var. graminis ALP genes are 96% and 88% homologous to the avenacinase gene from G. graminis var. avenae (Figure 5 and Table 4) , while the predicted amino acid sequences of the G. graminis var. tri tici and var. graminis ALPs are 96% and 91% similar, and 93% and 85% identical respectively (Figure 6 and Table 4) .

Preliminary screening of BamHl-digested genomic DNA of other fungi by Southern blot analysis using avenacinase or tomatinase DNA as a probe• revealed specific cross-hybridising sequences in the following organisms.

Tomato-attacking fungi:

Al ternari solani

Fusarium oxysporum f.sp. lycopersici Septoria lycopersici Bo try t is cinerea

Oat-attacking fungi:

Fusarium avenaceum Fusarium graminearum Fusarium culmorum Septoria avenae

Wheat/rye/grass-attacking:

Septoria nodorum Septoria tri tici

Pseudocercosporella herpa trichoideε Microdochium nival e Magnaporthe grisea Rhizoctonia solania

Other fungi:

Aspergillus nidulans Gibberella fujikuroi Fusarium solani

Nothing in Neurospora crassa .

The avenacinase and tomatinase nucleic acid sequences have been used to isolate cross-hybridising DNA sequences from genomic DNA libraries of two other fungal pathogens of tomato, Botrytis cinerea and Fusarium oxysporum f . sp. lycopersici . Both of these fungi produce enzymes which hydrolyse α-tomatine (Table 2) .

Genomic DNA clones which hybridise with avenacinase cDNA have also been isolated from a genomic DNA library of Magnaporthe grisea, a fungus for which saponin detoxification has not been described.

Antisera raised against avenacinase specifically recognise a single protein species in crude protein preparations of culture filtrates of the tomato pathogen Al ternaria solani , which is also known to degrade α-tomatine (Table 2) . EXAMPLE 4

ISOLATION OF RANDOM PEPTIDES WHICH INHIBIT SEPTORIA LYCOPERSICI TOMATINASE

Random peptides can be encoded into E. coli bacteriophage Fd gene III so that the peptide is presented on the phage surface. (See WO92/01047, for example, for details of the technology.) Such phage can be screened for binding to proteins. Large numbers of hexamers (for example) can be screened in this way and selected phage can be reamplified in E. coli and re-screened allowing selection of a few binding peptides from a very large starting population. As with antibodies a subset of this binding population may inhibit enzyme activity through competitive binding at the active site or through other allosteric effects.

A 5xl0⁸ clone random hexamer library was screened against purified tomatinase. Purified protein was biotinylated (NHS-LC biotin, Pierce Inc) and bound to streptavidin-coated 3 cm Petri dishes. 5xl0¹⁰ phage were panned against the immobliised protein and then washed to remove unbound phage. Binding phage were eluted with 0.1 M HCl and reamplified by infection into E coli and overnight growth. After four rounds of selection and re-amplification using successively less tomatinase (10 μg; 5 μg; 1 μg; 0.1 μg) , 100 phage clones were picked and their hexamer inserts sequenced.

The 100 phage clones represented a population of 82 different hexapeptides . ELISA tests indicated that 60% of these phage bound to tomatinase.

Tests were carried out to determine whether any of these binding phage inhibited the action of tomatinase, and inhibition was demonstrated for some phage (Figure 8) . The assays were carried out against the S-glucosidase substrate methyl umbelliferyl glucoside in the presence of a molar equivalent of phage (relative to tomatinase) and excess substrate.

It can be seen from Figure 8 that a number of the peptides on phage significantly inhibit tomatinase activity with one phage apparently increasing activity. Peptides were synthesised based on hexamer motifs in inhibitory phage, activator phage and representative binding phage which did not alter tomatinase activity. The amino acid sequences of these were QMAGLP, SVRTAY, LSEAAD, LPRKSH, EKRSIT, DYGFSR, GYGFSR and VRIRGS (see Figure 8 for their effect on tomatinase activity) . Preliminary enzyme assays with the synthetic peptides indicate that the same trend of inhibition is seen with the synthetic peptides as is seen with the hexamers expressed at the surface of the bacteriophage.

All documents mentioned herein are incorporated by reference. TABLE 1

SAPONINS IN PLANTS USED AS FOODS AND FEEDING STUFFS

Plants used as human foods:

Soya

Beans ( Phaseolus species)

Other beans and peas

Oats

Solanum species

Tomato

Allium species

Asparagus

Tea

Peanut

Spinach

Cucurbits

Sugar beet

Yam

Blackberry

Plants used as animal feeding stuffs

Alfalfa

Forage and cover crops

Sunflower

Horse chestnut

Guar

Lupin

Plants used as flavourings, health foods, tonics, etc:

Fenugreek

Liquorice

Nutmeg

Quillaja

Saponaria

Yucca

Gypsophila

Herbs

Edible seeds

Health foods

Ginseng TABLE 2

FUNGI AND PLANTS WHICH HAVE BEEN REPORTED TO HAVE SAPONIN GLYCOSYL HYDROLASE ACTIVITY (REFERENCES INDICATED IN BRACKETS) .

Saponin Fungi Avenacin Gaeumannomyces graminis var. avenae (15) Fusarium avenaceum (8)

Avenacosides Helminthosporium avenae (8) Septoria avenae

Tomatine Septoria lycopersici (16) Verticillium albo-atrum (18) Fusarium oxysporum f . sp. lycopersici (17) Botrytis cinerca (19) Al ternaria solani (8)

Plants

Tomatine Lycopersicon esculenmm (12) Solanine Phytophthora infestans (24) Fusarium caerulum (25)

Soyasaponin Aspergillus oryzae (23)

TABLE 3

N.T. : Not Tested ^a As determined by SDS-PAGE ^b The enzyme assays were carried out in the presence of excess substrate and at a pH at which both avenacinase and tomatinase can function effectively. ^c Partially purified enzyme fraction

Note: /3-glucosidase 1 and 2 from Gga do not hydrolyse saponin glycosides, do not cross-react with antiserum and do not hybridise with avenacinase or tomatinase c-DNA probes under stringent conditions. S-glucosidase 2 from Gga has only been partially purified so no specific enzyme activity could be calculated. However, the characterisation does show that /3-glucosidase 1 and 2 are different enzymes and are not identified, purified or cloned when following the subject method.

TABLE 4

ALP (Gg var tritici) ALP (Gg vargraminis) Tomatinase [S. lycopersici)

% % % % % % similar identical similar identical similar identical

Avenacinase DNA 95.9 88.5 60.3

AA 95.7 92.9 90.9 84.7 68.0 53.4

ALP (Gg var tritici) DNA 89.6 60.9

AA 92.8 87.4 69.1 54.6

ALP (Gg vargraminis) DNA 60.5

AA 69.3 54.1

Nucleotide (DNA) and Amino Acid (AA) relatedness of avenacinase, tomatinase and avenacinase like proteins.

REFERENCES

1) Price KR et al . (1987) . CRC Crit . Rev. Food Sci . Nutr. 26:27.

2) Mahato SB (1982) Phytochem. 21:959. 3) Steel CC and Drysdale RB (1988) . Phytochem. 27:1025.

4) Bangham AD et al . (1962) . Nature 196:952.

5) Dourmaskin RR et al . (1962) . Nature 194:1116.

6) Mahato SB et al . (1988) . Phytochem. 27:3037. 7) Chandel RS and Rastogi RP (1980) . Phytochem.

19:1889. 8) Schonbeck F and Scholsser E (1976) . Physiol . Plant

Pathol. (Ed. R Heitefuss and PH Williams)

Springer-Verlag, Berlin, pp653-678. 9) Mahato SB and Nandy AK (1991) . Phytochem. 30:1357 10) Tschesche R. Pharmacognosy and Phytochemistry, 1st

Int. Congress Munich 1970 (Eds H Wagner and L

Horhammer) , Springer-Verlag, Berlin, 1971, pp274-288. 11) Luning HU and Schlosser E (1975) . Z.

Pflanzenkrankh. Pflanzenschutz 82:699.

12) Prokoshev SM et al . (1956) Dokl . Akad. Nauk. SSR 106:313.

13) Henrissat B (1991) Biochem J 280:309. 14) Henrissat B and Bairach A (1993) Biochem J 293:781.

15) Turner EM (1961) J Exp Bot 12:169.

16) Durbin RD and Uchytil TB (1969) Biochim. Biophys. Acta 191: 176.

17) Ford JE et al . (1977) . Phytochem. 16:545. 18) Pegg GF and Woodward S (1986) . Physiol. Mol. Plant Pathol. 28:187.

19) Verhoeff K and Liem Jl (1975) . Phytopath. Z. 82:333.

20) Machida M et al. (1988) . Appl . Env. Microbiol. 54:3147.

21) Iijima M et al. Proc. Int. Symp. On New Technology of Vegetable Proteins, Oils and Starch Processing, Beijing, China, 1987, pp2-107.

22) Okubo K et al. (1986) . Chem. Senses 11:273. 23) Kudou S et al. (1990) . Agric. Biol . Chem. 54:131

24) Holland HL and Taylor GJ (1979) . Phytochem. 18:437.

25) McKee RK (1959) . J. gen. Microbiol. 20:681.

26) Jones et al (1992) Transgenic Research 1, 285-297 27) O'Sullivan and O'Gara, (1992) . Microbiological

Reviews. 56: 662-676 28) Thomashow, Weller, Bonsall and Pierson, (1990) . Applied and Environmental Microbiology. 56,4:000-000. 29) Hamdan, Weller and Thomashow, (1991) . Applied and Environmental Microbiology. 57,11:3270-3277.

30) Thomashow and Weller, (1988) . Journal of Bacteriology. 170,8:3499-3508.

31) Speakman and Kruger, (1984) . Journal of Plant Diseases and Protection. 91,4:391-395.

32) Capper and Higgins, (1993) . Plant Pathology. 42 :560-567

Claims

1. An isolated polypeptide having saponin glycosyl hydrolase activity.

2. A polypeptide according to claim 1 comprising a sequence of amino acids shown in Figure 4.

3. A polypeptide comprising an amino acid sequence which comprises an allele, derivative or mutant, by way of addition, insertion, deletion or substitution of one or more amino acids, of a sequence of amino acids shown in Figure 4.

4. A polypeptide according to claim 3 which has saponin glycosyl hydrolase activity.

5. An isolated polypeptide having saponin glycosyl hydrolase activity, which polypeptide has an amino acid sequence which is homologous to a sequence of amino acids shown in Figure 4.

6. A polypeptide comprising an amino acid sequence which comprises an allele, derivative or mutant, by way of addition, insertion, deletion or substitution of one or more amino acids, of a polypeptide according to claim 5.

7. A polypeptide according to claim 6 which has saponin glycosyl hydrolase activity.

8. A method of deglycosylating a saponin comprising contacting the saponin with a polypeptide according to any one of claims 1, 2, 4, 5 and 7.

9. A deglycosylated saponin obtainable using a method according to claim 8.

10. A nucleic acid isolate comprising a sequence of nucleotides encoding a polypeptide which comprises a sequence of amino acids shown in Figure 4.

11. Nucleic acid according to claim 10 wherein the sequence of nucleotides comprises any encoding sequence shown in Figure 4.

12. Nucleic acid according to claim 10 wherein the sequence of nucleotides comprises an allele, derivative or mutant, by way of addition, -insertion, deletion or substitution of one or more nucleotides, of any encoding sequence shown in Figure 4.

13. Nucleic acid comprising a sequence of nucleotides encoding a polypeptide, the polypeptide comprising an amino acid sequence which comprises an allele, derivative or mutant, by way of addition, insertion, deletion or substitution of one or more amino acids, of an amino acid sequence shown in Figure 4.

14. Nucleic acid according to any one of claims 10 to 13 wherein said polypeptide has saponin glycosyl hydrolase activity.

15. A nucleic acid isolate comprising a sequence of nucleotides encoding a polypeptide with saponin glycosyl hydrolase activity, said sequence comprising a nucleotide sequence complementary to a nucleotide sequence hybridisable with any encoding sequence shown in Figure 4.

16. Nucleic acid which is a vector comprising nucleic acid according to any one of claims 10 to 15.

17. Nucleic acid according to claim 16 further comprising regulatory sequences for expression of said polypeptide.

18. A method of making a polypeptide, which comprises expression from nucleic acid according to any one of claims 10 to 17.

19. A host cell comprising nucleic acid according to any one of claims 9 to 17.

20. A host cell according to claim 19 which is plant or microbial .

21. A method of producing a polypeptide, which comprises growing a host cell according to claim 19 or claim 20 under conditions for expression of said polypeptide.

22. A method according to claim 21 comprising isolation of said polypeptide following said expression.

23. A method of deglycosylating a saponin which comprises contacting the saponin with a polypeptide having saponin glycosyl hydrolase activity following production of the polypeptide in accordance with claim 18, claim 21 or claim 22.

24. A nucleic acid isolate comprising a sequence of nucleotides complementary to a nucleotide sequence encoding an amino acid sequence shown in Figure 4, or an allele, derivative or mutant, by way of addition, insertion, deletion or substitution of one or more amino acids, of an amino acid sequence shown in Figure 4.

25. Nucleic acid according to claim 24 wherein the sequence of nucleotides is complementary to an encoding nucleotide sequence shown in Figure 4.

26. Nucleic acid according to claim 24 wherein the sequence of nucleotides is complementary to a nucleotide sequence which is an allele, derivative or mutant, by way of addition, insertion, deletion or substitution of one or more nucleotides, of an encoding nucleotide sequence shown in Figure 4.

27. Nucleic acid according to any one of claims 24 to 26 which is DNA and which comprises regulatory sequences for expression from said sequence of nucleotides.

28. A host cell comprising nucleic acid according to any one of claims 24 to 27.

29. A host cell according to claim 28 which is plant or microbial.

30. A plant or any part thereof comprising a host cell according to claim 28 or claim 29.

31. Seed, selfed or hybrid progeny, or a clone or descendant of a plant according to claim 30, or any part thereof.

32. A method which comprises introduction of nucleic acid according to any one of claims 10 to 17, 24 to 27 into a host cell.

33. A method according to claim 32 wherein the host cell is a plant or microbial cell.

34. An oligonucleotide comprising a sequence encoding an amino acid sequence conserved between saponin glycosyl hydrolases or comprising a sequence complementary to a nucleotide sequence encoding a said amino acid sequence.

35. An oligonucleotide according to claim 34 comprising any of the sequences:

(i) GG.G/T) (A/C)CGC(C/T)ACCCA(T/C) (T/G)C(A/C)GGA;

(ii) TCGGC(T/C)TGCCA(A/C) (A/G) C(A/G)T(A/C) (T/OGGGTCC.

36. An oligonucleotide comprising a sequence encoding an amino acid sequence ATH(S/A)GV.

37. An oligonucleotide comprising a sequence complementary to a nucleotide sequence encoding an amino acid sequence DP(D/Y) (V/A)WQAD.

38. An oligonucleotide which comprises a sequence which is a variant or derivative, by way of addition, insertion, deletion or substitution of one of more nucleotides, of the sequence of an oligonucleotide according to any one of claims 34 to 37.

39. A method of obtaining nucleic acid encoding a polypeptide according to any one of claims 1 to 7 comprising hybridisation of an oligonucleotide according to claim 28, or a nucleic acid molecule comprising a said oligonucleotide, to target nucleic acid.

40. A method of obtaining nucleic acid encoding a polypeptide according to any one of claims 1 to 7, comprising hybridising to candidate nucleic acid a nucleic acid molecule which comprises (i) a nucleotide sequence encoding a polypeptide shown in Figure 4; (ii) a nucleotide sequence complementary to a nucleotide sequence encoding a polypeptide shown in Figure 4; (iii) a nucleotide sequence which comprises a nucleotide sequence encoding a polypeptide which is an allele, derivative or mutant, by way of addition, insertion, deletion or substitution of one or more amino acids, of any amino acid sequence shown in Figure 4, or a nucleotide sequence complementary thereto; (iv) any nucleotide sequence shown in Figure 4; (v) a nucleotide sequence complementary to any nucleotide sequence shown in Figure 4; (vi) a nucleotide sequence which comprises an allele, derivative or mutant, by way of addition, insertion, deletion or substitution of one or more nucleotides, of any nucleotide sequence shown in Figure 4, or a nucleotide sequence complementary thereto; or (vii) a nucleotide sequence which is a fragment of any of (i) - (vi) .

41. A method according to claim 39 or claim 40 wherein the hybridisation is followed by identification of successful hybridisation and isolation of target nucleic acid.

42. A method according to any one of claims 39 to 41 involving use of nucleic acid amplification.

43. A method of obtaining an antibody or a fragment thereof able to bind a saponin glycosyl hydrolase, comprising immunising a mammal with a polypeptide according to any one of claims 1 to 7 or a fragment thereof and isolating an antibody or fragment thereof able to bind a saponin glycosyl hydrolase.

44. A method of obtaining an antibody or a fragment thereof able to bind a saponin glycosyl hydrolase, comprising screening an expression library of polypeptides comprising antibody variable domains for binding to a polypeptide according to any one of claims 1 to 7 or a fragment thereof.

45. A method of obtaining a polypeptide with saponin glycosyl hydrolase activity comprising, following obtaining an antibody or a fragment thereof using a method according to claim 43 or claim 44, screening candidate polypeptides with the antibody or fragment for binding thereto.

46. A method according to claim 45 wherein the antibody or fragment thereof is able to bind avenacinase.

47. A method according to claim 45 or claim 46 wherein a polypeptide bound by the antibody or fragment thereof is isolated.

48. A method according to claim 47 wherein the isolated polypeptide is at least partially sequenced and the amino acid information so obtained is used in obtaining nucleic acid encoding the polypeptide.

49. Use of an antibody able to bind a polypeptide with saponin glycosyl hydrolase activity in obtaining a different polypeptide with saponin glycosyl hydrolase activity.

50. A method of screening for a molecule able to modulate activity of a saponin glycosyl hydrolase, comprising contacting a candidate molecule with a polypeptide according to any one of claims 1 to 7 and determining saponin glycosyl hydrolase activity.

51. A molecule able to modulate activity of a saponin glycosyl hydrolase.

52. A molecule according to claim 51 which is a peptide or a polypeptide.

53. A molecule according to claim 52 which comprises an amino acid sequence shown in Figure 8 or a variant or derivative thereof, by way of addition, insertion, deletion or substitution of one or more amino acids.

54. A method which comprises expression of a molecule according to claim 52 or 53 from encoding nucleic acid therefor.

55. A method according to claim 54 wherein the expression takes place in a host cell.

56. A method according to claim 55 wherein the host cell is plant or microbial.

57. A method of modulating activity of a saponin glycosyl hydrolase, comprising contacting the saponin glycosyl hydrolase with a molecule according to any one of claims 51 to 53.

58. A method of modulating activity of a saponin glycosyl hydrolase, comprising contacting the saponin glycosyl hydrolase with a molecule according to claim 52 or claim 53 following expression of the molecule using a method according to any one of claims 54 to 56.

59. A host cell containing nucleic acid encoding a molecule according to claim 52 or claim 53.

60. A host cell according to claim 59 which is plant or microbial.